Gpu-boosted software tool emulates wireless signals in real time

MIRAGE: A GPU-Accelerated IQ-Level Channel Emulator for Real-Time Wireless Digital Twins

Networking and Internet Architecture

Summary

Wireless networks need to be tested before they are used in real life, but simulating how signals travel and get distorted is tricky and slow. The authors created MIRAGE, a program that uses powerful graphics processors (GPUs) to quickly mimic these wireless signal effects on digital data streams for 5G networks. This tool works in real time and can handle many users and antennas at once without losing accuracy. Their tests show it can run fast enough to keep up with real network timing and help plan wireless setups before using physical equipment.

What this means in practice

  • For wireless network engineers: Test and evaluate wireless network policies in real time using a GPU-accelerated channel emulator before actual deployment.
  • For telecom operators: Simulate and optimize resource allocations and network slices to meet throughput targets under various signal conditions before physical changes.

Authors

Ali Saeizadeh, Angelo Feraudo, Avaneeth Anil, Leonardo Bonati, Pedram Johari, Tommaso Melodia

Abstract

Wireless digital twins enable network policies to be evaluated before deployment, but reproducing channel effects within a running protocol stack requires continuous waveform processing under strict timing constraints. We present MIRAGE, a GPU-accelerated channel emulator that applies propagation effects directly to in-phase and quadrature samples exchanged by software 5G New Radio stacks. MIRAGE combines batched multiuser, multiantenna filtering with timestamped shared-memory transport and radio-driver integrations for OpenAirInterface and OCUDU. External propagation models supply channel responses asynchronously, while versioned buffers ensure consistent coefficients within each processing interval. Experiments on one NVIDIA H100 GPU evaluate waveform accuracy, processing time, channel updates, and full-stack traffic delivery. At 61.44 million samples per second, eight single-antenna links, four two-by-two links, and one four-by-four link meet the tested 0.5 ms bidirectional processing criterion. Waveform errors remain within one integer quantization step per component, with no mixed coefficient versions across 16.7 million analyzed slots. Integration with the X5G testbed enables slice resource allocations to be evaluated before deployment. Both evaluated slices ultimately satisfy their throughput caps, requiring one physical reconfiguration under line-of-sight conditions and four under non-line-of-sight conditions. These results demonstrate full-stack policy evaluation and expose prediction errors affecting transfer to a physical network.